With interest we read the article by Kerr et al. about the diagnostic work-up of 36 patients with confirmed mitochondrial disorder(MID) (MID-group) and 80 patients with suspected MID but yet unconfirmed diagnosis (NO-group) by means of next generation sequencing(NGS).[1] In the MID-group MID was diagnosed upon a traditional approach.[1] The authors found that bigenomic sequencing(BGS) confirmed the diagnosis in 13/36 MID-patients and detected new pathogenic variants in 3 other MID-patients. In the NO-group one new MID-patient was identified by NGS.[1] We have the following concerns.

We do not agree with the conclusions that bigenomic sequencing(BGS), consisting of nWES and optimised mtDNA analysis, should be the first step in investigating suspected MIDs.[1] BGS confirmed an already known diagnosis in only 13/36 patients, and provided a new diagnosis in only 1/80 patients with suspected MID. How do the authors explain that 20 patients with a definite MID had a negative genetic test. This is unusual particularly as the diagnosis MID was supported by different methods. 

Furthermore, we should know why only 36 of the 115 patients with confirmed MID underwent NGS. Was this due to pecuniary reasons, due to absent consent, or unavailability of invited patients? 

Of the 390 patients referred for suspected MID according to figure 1, 69 had another diagnosis according to table 1. However, “unclassified neurodegenerative disease” is no “other” diagnosis. Furthermore, it is surprising that sepsis, juvenile osteoporosis, and bronchogenic lung carcinoma, as listed in table 2, initially suggested MID. 

In table 2 the authors use the term “bilateral bulbar neurosis” [1]. We assume that this in an error and actually should be “bilateral retrobulbar optic neuritis”.

Overall, diagnosing MIDs should be based not only on NGS but also on traditional procedures. NGS not only provides false positive and false negative results but still has a low sensitivity.